In this project, we will develop an understanding of the range of clinical symptoms and biological factors (genetics and brain morphometry) that correlate with obsessive-compulsive behavior in autism. Autism is a highly heterogeneous disorder. A significant number of patients do not improve substantially with current treatments. We refer to this group of patients as difficult-to-treat autism (DTT-Autism). Our preliminary data suggest that these patients exhibit obsessive-compulsive behavior (OCB). We will test the central hypothesis that participants with autism with high OCB represent a subtype of difficult-to-treat autism (DTT-Autism) who will have abnormal maturation of frontal-striatal circuitry and genetic susceptibilities analogous to those previously studied in obsessive-compulsive spectrum (OCS) conditions. Capitalizing on recent progress in neuroimaging and genetics in OC spectrum (OCS) disorders, this project will test the hypothesis that autism with OCB will share genetic and brain circuitry changes that have been demonstrated in OCS disorders. We will test the hypothesis that autism with OCB will correlate with abnormalities in frontal-striatal circuitry as is true for OCD and related OCS disorders. We will also look for associations between rare and common variation in OC-related genes and OCB symptoms in autism. Using a combination of approaches including clinical assessment, neuroimaging and genotyping, we will characterize the subtype of autism with OCB. This work is important as it studies a group of autism patients who are in greatest need of treatment development. If our hypotheses about the strong biologic relationship between OC spectrum disorders and autism with OCB are accurate, novel treatments for autism may be drawn from ongoing research in interventions in treatment-refractory OCD.